Hypertension Following Coronary Artery Bypass Surgery

Role of Preoperative Propranolol Therapy

PAUL K. WHELTON, M.D., JOHN T. FLAHERTY, M.D., NIALL P. MACALLISTER, M.D.,

LEVI WATKINS, M.D., A L A N POTTER, B.A., DOLORES JOHNSON, B.A.,

R. PATTERSON RUSSELL, M.D., AND W. GORDON WALKER, M.D.

SUMMARY Over a 9-month period, the incidence and characteristics of hypertension followtng coronaryartery bypass surgery were studied in a group of 52 patients. Hypertension occurred in 61% of the patients andwas characterized by an increase in arterial blood pressure of 35 ± 2 mm Hg mean ± SEM during the earlypostoperative period. Preoperative blood pressures and hemodynamk variables were similar in those whodeveloped hypertension and those who remained normotenshe. Ninety-four percent of those who developedhypertension as compared to only 40% of those who remained normotensive received propranolol during the 24hours preceding surgery (x1 = 15.4; p < 0.001). Maximal blood pressures during the first 5 hours following thetermination of cardiopulmonary bypass were significantly positively correlated with preoperative propranololdosage (p < 0.01). Hypertension was not associated with significant changes in plasma renin activity orangiotensin II levels, but concomitant plasma catecholamine concentrations were elevated significantly (p <0.005). However, a similar rise in plasma catecholamine concentrations was found in those who remained nor-motensive. Hypertension was associated with an increase in systemic vascular resistance (p < 0.001) and leftventricular stroke work index [p < 0.05), and a fall in stroke volume (p < 0.005) and cardiac index (p <0.001). These studies suggest that hypertension following coronary artery bypass surgery is common, resultsfrom an increase in systemic vascular resistance, is not renin-angiotensin mediated, and may, in part, berelated to preoperative propranolol administration. (Hypertension 2: 291-298, 1980)

KEY WORDS • hypertension • coronary artery bypass surgery • renin activitycatecholamine • hemodynamics

DURING a 3-month period from March toMay, 1977, more than 70% of patientsundergoing coronary artery bypass surgery

at The Johns Hopkins Hospital experiencedvasomotor instability within the first 24 hours follow-ing revascularization. While hypotension requiring theadministration of vasopressor agents was occasion-ally noted, the more common problem was that ofhypertension. Since the incidence and severity ofpostoperative hypertension noted in this group wasboth surprising and disturbing, the circumstances sur-rounding the development of hypertension werecarefully examined. In particular, we were anxious todefine the hypertensive patients' clinical character-istics and to assess the roles of the sympathetic andrenin-angiotensin systems in the genesis of this formof hypertension.

From the O'Neill Research Laboratories of the Department ofMedicine and the Departments of Anesthesia and Surgery, TheJohns Hopkins University School of Medicine, Baltimore,Maryland.

Supported in part by USPHS granU HL 03303, RR 35, AM05655, NHLI KO4HL00O19-01, and The Jane Hilder Harris Foun-dation.

Address for reprints: Paul K. Whelton, M.D., Division ofNephrology, The Johns Hopkins Hospital, Baltimore, Maryland21205.

Received May 25, 1979; accepted January 8, 1980.

Methods

During a 9-month period from June, 1977, toMarch, 1978, 52 of 204 patients who underwent cor-onary artery bypass surgery at The Johns HopkinsHospital participated in these studies. These 52 wereall carefully monitored through the postoperativeperiod by one of us (AP). Patients with knownhypertension or a past history of hypertension wereexcluded, but the study group was otherwise un-selected. In every patient, preoperative functional car-diac capacity was evaluated by means of the NewYork Heart Association criteria.1 Angiographic cor-onary artery disease was defined as the presence of avessel lumen narrowing equal to or greater than 50%.Preoperative hemodynamic variables measured at thetime of cardiac catheterization included mean arterialblood pressure (MAP) in mm Hg, left ventricular enddiastolic pressure (LVEDP) in mm Hg, left ven-tricular end diastolic volume (LVEDV) in ml, and leftventricular ejection fraction (LVEF) in %.

Standard anesthetic and surgical techniques wereutilized in every patient. Preanesthetic medications in-cluded morphine and diazepam. Endotracheal intuba-tion was facilitated by means of pancuroniumbromide, and all patients were mechanically ventilatedthroughout the operative period. Anesthesia was in-duced by intravenous diazepam, morphine, or

291

by guest on July 19, 2018http://hyper.ahajournals.org/

Dow

nloaded from

292 HYPERTENSION VOL 2, No 3, MAY-JUNE 1980

Abbreviations Used

AH = angiotensln IICAT = catecholamine concentration

CI = cardiac indexCPB = cardiopulmonary bypassCVP = central venous pressure

LVEF = left ventricular ejection fractionLVEDP = left ventricular end diastolic pressureLVEDV = left ventricular end diastolic volumeLVSWI = left ventricular stroke work index

MAP = mean arterial blood pressureMPAP = mean pulmonary artery blood pressurePCWP = pulmonary capillary wedge pressure

PRA = plasma renin activityPVRI = pulmonary vascular resistance index

SV = stroke volumeSVRI = stroke volume resistance index

thiamylol sodium. Subsequent anesthesia was main-tained by fentanyl, morphine, or halothane, in addi-tion to nitrous oxide and oxygen.

Hyperkalemic cardiac arrest was employed in themajority of cases, and standard cardiopulmonarybypass (CPB) with nonpulsatile flow and a bubbleoxygenator was utilized in all patients. Pump flowrates were maintained in a range of 2 to 2.5liters/min/m3. Mild systemic (27°-29° C) and inmany patients severe local myocardial (4° C) hypo-thermia was employed. The root of the aorta wascross-clamped either intermittently or continuouslyfor periods varying from 10 to 60 minutes.

All patients underwent coronary artery bypass withreversed saphenous vein aortocoronary grafts. Twopatients received prosthetic heart valves, and one un-derwent concomitant ventricular aneurysmectomy.Before bypass, each patient received heparin (4mg/kg), which was subsequently counteracted by ad-ministration of protamine sulphate. Blood and intra-venous fluids were administered when indicated bychanges in hematocrit, systemic arterial bloodpressure, and left ventricular filling pressure.

Following surgery, all patients were transferred tothe surgical intensive care unit where artificial ventila-tion was continued and vital signs and blood gasesmonitored. Postoperative pain and/or anxiety werealleviated by use of morphine or diazepam. Cardiacoutputs were determined by thermodilution tech-niques, and systemic arterial blood pressure by meansof a radial artery cannula connected to a Statham P23Db pressure transducer. The MAP was obtainedeither electrically or by adding to the diastolic bloodpressure one-third of the pulse pressure. Postoperativehypertension was defined as the occurrence of a persis-tent (greater than 15 minutes) elevation of MAP tolevels greater than 105 mm Hg during the first 5 hoursfollowing the termination of CPB.

In 24 of the 52 patients, perioperative hormonalsampling studies were performed. Blood samples froman indwelling arterial catheter were obtained im-mediately following the institution of generalanesthesia, immediately following and again 15minutes following the discontinuation of CPB, andimmediately before treatment of a hypertensiveepisode, or 300 minutes following the termination ofCPB if hypertension did not occur. Samples for deter-mination of plasma renin activity (PRA) in ng/ml/ hrand angiotensin II (All) in pg/ml were obtained andassayed as previously described,' while those foranalysis of total plasma catecholamine concentration(CAT) in pg/ml were obtained and assayed by amodification of the method of Passon and Peuler.1

In 19 of the 52 patients, serial hemodynamic andbody temperature measurements were obtainedfollowing the termination of CPB. In each patient,control values were taken as those obtained during thefirst 30 minutes. In those who developed hypertension,control observations were compared to measurementsobtained just prior to the treatment of hypertension.In patients who remained normotensive, controlresults were compared to a mean of those obtainedbetween 60 and 180 minutes following the terminationof CPB. Body temperature was recorded by means ofa rectal probe. The hemodynamic measurements ob-tained included MAP, heart rate in beats/min, centralvenous pressure (CVP) in mm Hg, pulmonarycapillary wedge pressure (PCWP) in mm Hg, meanpulmonary artery blood pressure (MPAP) in mm Hg,stroke volume (SV) in ml, and cardiac index (CI) inliters/min/m2. The following formulas were used tocalculate systemic vascular resistance index (SVRI) inunits/m3, pulmonary vascular resistance index(PVRI) in units/m1, and left ventricular stroke workindex (LVSWI) in g/m/m1:

SVRI =

PVRI =

MAP -CVPCI

MPAP-PCWPRI

LVSWI = SV (MAP - PCWP) X 0.0136.

Group data are expressed as the mean value ± stan-dard error of the mean (SEM). Statistical proceduresused to evaluate the significance of observeddifferences included the chi square test, with a Yatescorrection, to compare nonquantitative results, andthe Student's t test to compare results when thevariable was a measured value. Correlations betweenvariables were analyzed by means of bivariate andmultivariate linear regression analysis.

Results

The clinical and laboratory characteristics of the 52patients are outlined in table 1. Thirty-two (61%)developed hypertension, while the remaining 20 (38%)remained normotensive. The majority of both thehypertensive and normotensive patients were middle-aged white men with severe angina and angiographic

by guest on July 19, 2018http://hyper.ahajournals.org/

Dow

nloaded from

HYPERTENSION FOLLOWING CORONARY SURGERY/ Whelton et al. 293

TABLE 1. Preoperative Clinical and Laboratory Character-istics of 38 Patients who Developed Hypertension and SO whoRemained Normotensive Following Coronary Artery BypassSurgery

Hypertensive Normotenaivepatients patientsPatient characteristics

No. 32 20

Age (yrs)

Race: % white

Sex: % male

Prior MI (%)

Functional capacity:

Class IV (%)

aass in (%)Class II (%)

CA narrowing > 50% atcoronary angiography:

3 or more (%)

2 CA (%)

1 CA (%)

left main (%)

left anterior descending

left circumflex (%)

riirht (%)

54 ± 1*

94

81

48

44

34

22

69

22

9

25

87

72

78

53 ± 2*

100

90

31

42

42

16

53

31

16

26

79

58

74

Hemodynamic parameters at cardiac catheterization:

MAP (mm Hg) 88 * 3 87 ± 3

LVEDP (mm Hg) 12 •*= 1 10 =*> 1

LVEDV (ml) 153 ± 9 133 * 17

LVEF (%) 63 ± 2 60 ± 3

•Values represent mean ± SEM.MI =• myocardial infarction; CA = coronary artery;

MAP — mean arterial blood pressure; LVEDP = left ven-tricular end diastolic pressure; LVEDV = left ventricularend diastolic volume; LVEF = left ventricular ejectionfraction.

evidence of significant disease in all three major cor-onary arteries. Furthermore, the location of coronaryartery disease and results of the preoperativehemodynamic measurements were similar in bothhypertensive and normotensive patients.

There was a significant difference, however, in thefrequency of preoperative propranolol administrationin both groups (table 2). Thirty (94%) of the 32hypertensive patients as compared to only eight (40%)of the 20 normotensive patients received propranololduring the 24 hours preceding surgery (xl = 15.4; p <0.001). In addition, the eight normotensive patientsreceived significantly less propranolol (56 ± 10 mg)(p < 0.001) than their 30 hypertensive counterparts(129 ± 18 mg). Heart rate immediately prior to in-duction of general anesthesia was also significantlylower in the hypertensive than normotensive group{p < 0.005). As shown in table 3 a strong positive cor-

relation existed between preoperative propranololdosage and maximal MAP during the first 5 hoursfollowing termination of CPB. Maximal MAP did notcorrelate significantly with age, however, nor withpreoperative functional cardiac capacity, coronaryangiography, or hemodynamic variables measured atcardiac catheterization.

Although there was some variation in anestheticand surgical techniques, there was no pattern of usethat allowed one to differentiate hypertensive fromnormotensive patients. Similarly, the two groups ex-hibited no significant difference in the duration ofCPB or the complexity of the surgery as judged by thenumber of patients requiring additional heart surgery.

Thirty-one of the 32 patients who developedhypertension did so within 2 hours of the terminationof CPB, and 14 (44%) of them developed it within thefirst of these 2 hours. Every hypertensive patient had aMAP equal to or greater than 110 mm Hg, and 19(59%) had a MAP equal to or greater than 120 mmHg. In all patients, blood pressure was rapidly andeffectively lowered by continuous intravenous infusionof sodium nitroprusside. Based on the duration ofnitroprusside administration, hypertensive episodesappeared to last from 4 to 12 hours.

The perioperative blood pressure and hormonalcharacteristics of the 15 hypertensive and nine nor-motensive patients in whom hormonal samplingstudies were performed are outlined in table 4. Initialmeasurements of MAP, PRA, All, and CAT, ob-tained immediately following the induction of generalanesthesia, were similar in the two subgroups ofhypertensive and normotensive patients. During CPB,blood pressure was decreased (p < 0.001) and CATwas increased {p < 0.005) significantly but equally inboth subgroups, while levels of PRA and Allremained unchanged. At 15 minutes following the ter-mination of CPB, blood pressure had increasedtoward normal levels in both subgroups {p < 0.025),and this was accompanied by a modest fall in CAT butunchanged values of PRA and All. The subsequentoccurrence of hypertension was associated with nofurther significant changes in hormonal levels. Thus,while hypertension was associated with elevated levelsof CAT, these values were similar to those previouslyobtained during surgery and were no higher than thosemeasured 300 minutes following the termination ofCPB in patients who remained normotensive.

Significant positive correlations were noted betweenMAP and PRA during the first two sampling periods.These were apparent both in the entire group (r = +0.58, p < 0.005 for the first period; and r = + 0.57,p < 0.005 for the second period) and in the subgroupof 15 patients who subsequently developed hyperten-sion (first period, r = + 0.61, p < 0.02; and second pe-riod, r = + 0.71, p < 0.005). However, no significantcorrelations existed between maximal MAP and ab-solute hormonal values, or with the differences notedbetween values obtained at the differential samplingperiods. As was the case in the overall group, a strongpositive correlation existed between maximal MAPand preoperative propranolol administration (r =

by guest on July 19, 2018http://hyper.ahajournals.org/

Dow

nloaded from

294 HYPERTENSION VOL 2, No 3, MAY-JUNE 1980

TABLE 2. Operative Characteristics of the SS Patients who Developed Hypertension and the SO who RemainedNormotensive Following Coronary Artery Bypass Surgery

Patient characteristics

No.

Propranolol use before surgery:

3 days (%)

2 days (%)

1 day (%)

Propranolol dose before surgery:

3 days (mg/24 hr)

2 days (mg/24 hr)

1 day (mg/24 hr)

Heart rate before anesthesia (beats/min)

Bypass grafts/patient (no.)

Concomitant surgery:

Heart valve replacement (%)

Ventricular aneurysmectomy (%)

Duration CPB (min)

Peak MAP during 300 min after CPB(mm Hg)

Time of peak MAP after CPB (min)

Hypertensivepatients

32

94

94

94

194 ± 25f

185 ± 23

129 ± 18

68 ± 2

2.9 * 0.2

3

0

114 ± 7

121 =* 2

81 ± 6

P*

< 0.001

< 0.001

< 0.001

< 0.001

< 0.001

< 0.001

< 0.005

< 0.001

< 0.001

Normotensivepatients

20

45

45

40

124 ± 18f

103 ± 18

56 ± 10

79 ± 2

2.7 * 0.2

5

5

111 ± 7

88 ± 3

149 ± 14

•Values represent probability that observed difference is due to random sampling variation.fValues represent mean ± SBM.MAP = mean arterial blood pressure; CBP = cardiopulmonary bypass.

TABLE 3. Specific Bivariate Correlations for the Entire Group and for the Hypertensive Subgroups in whomHormonal Samples or Hemodynamic Observations were Obtained

Patient group Variables tested for correlationCorrelationcoefficient

Pvalue

All 52

15 hypertensives withhormonal sampling

11 hypertensives withhemodynamicobservations

Maximal

Maximal

Maximal

Maximal

Maximal

Maximal

Maximal

A MAPt

A MAP

A MAP

MAP* vs

MAP vs

MAP vs

MAP vs

MAP vs

MAP vs

MAP vs

vs

vs

vs

age

preoperative functional capacity

propranolol 3 days before surgery

propranolol 1 day before surgery

PRAfplasma A II concentration!

plasma catecholamine concentrationf

A

A LVSWIJ

A decrease in CIJ

-0.031-0.119

+0.401+0.352

+0.191

-0.205

+ 0.027

+0.635

+0.688+ 0.464

NSNS

< 0.003

< 0.01

NS

NS

NS

< 0.05

< 0.02

NS

•Maximal MAP = maximal mean arterial blood pressure during first 5 hours following termination ofcardiopulmonary bypass (CBP).

fSampling performed immediately prior to treatment of hypertensive episode. PRA = plasma renin activity;A II = angiotensin II.

{Difference between control and hypertensive observations. SVRI = systemic vascular resistance index;LVSWI = left ventricular stroke work index; CI = cardiac index.

by guest on July 19, 2018http://hyper.ahajournals.org/

Dow

nloaded from

HYPERTENSION FOLLOWING CORONARY SURGERY/Whelton et al. 295

TABLE 4. Comparison of Perioperalive Blood Pressure and Hormonal Characterisiics in 15 Patients who Developed PostoperativeHypertension (HT) and Nine who Remained Normotensive (NT)*

Sampling times

Immediately after inductionof general anesthesia

Immediately beforetermination of CPB

At 15 min aftertermination of CPB

Immediately before treatsment of HT episode

At 300 min aftertermination of CPB

MAP(mm Hg)

HTpts

87±3

64±3

84±3

122±3

NTpts

92±4

67±4

86±3

79±2

PRA(ng/ml/hr)

HTpts

0.51±0.09

1.01±0.23

0.71±0.27

0.53±0.21

NTpts

0.73±0.23

0.79±0.19

1.27±0.67

1.00±0.23

A l l(pg/ml)

HT pts

7.8±1.2

7.1±1.3

8.7±2.7

6.0±1.4

NTpts

9.4±2.7

7.3±1.3

10.0±3.7

10.5±1.8

Plasma catecholamine(pg/ml)

HTpts

347±64

1325±303

933±70

1132±143

NTpts

370±101

1620±280

895±182

1636±328

*Values represent mean ± BKM.CPB = cardiopulmonary bypass.

+ 0.50, p < 0.02 for propranolol dosage 3 days priorto surgery; and r = + 0.42, p < 0.05 for propranololdosage 1 day prior to surgery).

The relationship of factors other than preoperativepropranolol administration to maximal MAP duringthe first 5 hours following termination of CPB wasfurther evaluated by means of multiple linear regres-sion analysis. In this analysis, maximal MAP wastreated as the dependent variable, while age, func-tional cardiac capacity, preoperative propranololdosage 3 days prior to surgery, and hormonal valueswere treated as the independent variables. When themultiple correlation coefficient was calculated for

these independent variables, r was + 0.60 and p <0.002. Since the correlation between maximal MAPand propranolol dosage alone was r = 0.50, it is clearthat the contribution of preoperative propranolol ad-ministration to this relationship was greater than thesum of all the other variables tested.

Results of the hemodynamic and body temperaturemeasurements obtained in the 19 patients in whomserial observations were performed following the ter-mination of CPB are outlined in table 5. Controlmeasurements were similar in the 11 patients who sub-sequently developed hypertension and the eight whoremained normotensive. In the latter eight patients

TABLE 5. Results of Hemodynamic Studies Performed in 11 Patients who Developed Hypertension and Eightwho Remained Normolensive Following Termination of Cardiopulmonary Bypass (CPB)

Hemodynamic studies

Hypertensive patients*

Control

Beforehypertension

therapy

Normotensive patients*Mean between60 and 180 min

Control after end of CPB

Time from end of CPBto sampling (min) 22 ± 4

MAP (mm Hg) 82 ± 3

SVRI (unite/m1) 26.8 ± 2.4

CI (liters/min/m») 3.1 ± 0.2

SV (ml/beat) 64 ± 4

Heart rate (beats/min) 93 * 2

LVSWI (g/m/m1) 30.9 ± 1.4

PCWP (mm Hg) 12 ± 1

MPAP (mm Hg) 18 ± 2

PVRI (units/m1) 2.1 ± 0.4

Body temperature (° C) 36.2 ± 0.4

< 0.001

< 0.001

< 0.001

< 0.005

< 0.05

73

114

46.0

2.4

53

94

37.8

11

16

2.3

36.0

± 19

± 4

± 2.5

± 0.1

± 3

± 4

± 2.4

± 1

± 1

± 0.4

± 0.2

16 ± 4

85 ± 4

26.5 ± 2.3

2.6 ± 0.2

56 ± 5

90 ± 4

31.8 ± 2.0

10 ± 2

16 ± 2

2.4 ± 0.3

36.2 ± 0.4

107 ± 16

83 ± 1

28.6 ± 3.2

2.6 ± 0.2

56 ± 5

88 ± 5

29.3 ± 2.5

11 ± 1

18 ± 2

2.8 ± 0.7

36.7 ± 0.5

'Values represent mean ± SEM.fValues represent probability that observed difference is due to random sampling variation.CPB = cardiopulmonary bypass; CI = cardiac index; LVSWI = left ventricular stroke work index; MAP

= mean arterial blood pressure; MPAP = mean pulmonary artery blood pressure; PCWP = pulmonarycapillary wedge pressure; PVRI = pulmonary vascular resistence index; SV = stroke volume; SVRI =stroke volume resistence index.

by guest on July 19, 2018http://hyper.ahajournals.org/

Dow

nloaded from

296 HYPERTENSION VOL 2, No 3, MAY-JUNE 1980

who continued to be normotensivc, a mean ofmeasurements obtained between 60 and 180 minutesfollowing the termination of CPB showed results thatwere unchanged from control values. In the 11 whodeveloped hypertension, measurements obtained justprior to the treatment of the hypertensive episodedemonstrated a significant rise in SVRI {p < 0.001)and LVSWI (p < 0.05) and a significant fall in CI(p < 0.001) and SV (p < 0.005). As shown in table 3,the changes in SVRI and LVSWI were significantlycorrelated with concomitant changes in MAP (p <0.05 and p < 0.02 respectively). Hypertension was notassociated with any significant changes in heart rate,PCWP, MPAP, and PVRI, or with any significantchange in body temperature.

Discussion

Since the original report of Estafanous et al.4 in1973, it has become increasingly clear that hyperten-sion may occur following coronary revascularizationprocedures. Indeed, this constitutes one of the mostfrequent complications following this form of surgery;it has been reported to occur in 33% to 76% ofpatients.8'* In our present study, using the criteriapreviously outlined, hypertension was noted in morethan 60% of patients following successful surgery.Whatever the exact incidence, postoperative hyperten-sion is a common and potentially dangerous complica-tion, likely to increase morbidity from several causes.Certainly, elevations in blood pressure similar to thosenoted in our study are recognized as risk factors in thegenesis of acute cerebrovascular accidents,' especiallyin arteriosclerotic patients undergoing cardiacrevascularization. Since patients undergoing coronaryartery bypass surgery may have areas of cardiacischemia that cannot be bypassed, the risk ofprecipitating a myocardial infarction is significant.The increase in LVSWI noted in our study would beexpected to increase myocardial oxygen demands,which in the presence of a fixed coronary blood supplymight lead to regional ischemia. Laboratory datasuggestive of subendocardial hypoperfusion havepreviously been reported to occur in this situation.6 In-creased postoperative bleeding from suture lines hasalso been reported during hypertensive episodesfollowing coronary artery bypass surgery.*'10

The hypertension noted in our study bore manysimilarities to that previously reported by other in-vestigators; it occurred within the first few hoursfollowing surgery, appeared to result from an increasein peripheral vascular resistance,"1 '• '• n and was nor-malized by administration of intravenous sodiumnitroprusside.4-7> 10> 1J However, in contrast to thefindings of Roberts et al.7 and Fouad et al.,u our find-ings showed that hypertension was associated with asignificant fall in SV and CI and an increase inLVSWI. The importance of these observations is un-certain since concomitant measurements of PCWPwere unchanged from control values and within thenormal range. The finding of normal wedge pressures,however, may only reflect the fact that hypertension

was always treated as soon as our criteria fordiagnosis were satisfied. These studies have all beenconducted during a relatively unstable physiologicstate, and hence different sampling times may in partbe responsible for differences between the present andprevious studies.4"*1 *~12 It is of interest that increases inSV and CI following successful therapy of this form ofhypertension have been reported by Niarchos et al.18

but not Tarazi et al.14

Although investigators agree on the generalcharacteristics of hypertension and its associatedhemodynamic changes, they do not agree as to thepathophysiologic mechanisms involved in its produc-tion. They have attributed hypertension to activationof a coronary artery chemoreflex mechanism,7 apressor response to reflex activation of other cardiacor aortic receptors,14 a response to varying anestheticor operative techniques,10 or to postoperative pain andanxiety. Studies of the sympathetic and renin-angiotensin systems have revealed conflicting results,with some suggesting stimulation of adrenergic7'' andothers renin-angiotensin systems.6'7> 1B

Reflex cardiogenic hypertension has been reportedas a response to stimulation of both chemoreceptorsand mechanoreceptors. Although this type ofhypertension has most frequently been reportedfollowing stimuli arising within the distribution of theleft coronary artery, clearly mechanoreceptors exist inother areas of the heart and great vessels. Tarazi etal.14 have presented a strong case for the role of a sym-pathetically mediated hypertensive reflex followingcardiac surgery. Using unilateral stellate ganglionblockade, they achieved normalized blood pressure in18 (67%) of 27 patients who developed hypertensionfollowing coronary artery bypass surgery, and infer-entially suggested that the hypotensive response wasdue to interruption of the afferent limb of a sym-pathetic pressor reflex.

In our study, there is little evidence to support a rolefor a coronary artery hypertensive reflex.171" Bothhypertensive and normotensive patients had similarfunctional and anatomic coronary artery disease,similar surgical procedures as judged by the number ofbypass grafts performed, and similar durations ofCPB. In contrast to the findings of Roberts et al.,7 wefound that vascular lesions in the main left coronaryartery were equally common among our hypertensiveand normotensive patients. Although variation inanesthetic or operative techniques may influenceblood pressure control significantly and thus con-stitute an important mechanism in the development ofhypertension, we found no evidence to support thispossibility. The effects of postoperative pain and anx-iety are hard to evaluate, but in this and other studies,use of tranquilizers and analgesic agents have failed toprevent the occurrence of hypertension.4' "•7 Severalauthors'-7i 1S have suggested an important etiologicrole for the renin-angiotensin system, but we couldfind no evidence for such an association. In thisregard, our results agree with those of Wallach et al.'Most investigators have reported adrenergic stimula-tion during and after coronary artery bypass sur-

by guest on July 19, 2018http://hyper.ahajournals.org/

Dow

nloaded from

HYPERTENSION FOLLOWING CORONARY SURGERY/ Whelton et al. 297

gery.7' *• " • " This is not surprising, since the sym-pathetic nervous system is known to respond tostimuli accompanying cardiac surgery, includinganesthetic effects, hypothermia, hypotension (such asthat noted during CPB), hypoxia, hypercarbia, andanxiety or pain." Although we found that hyperten-sive patients had increased levels of circulatingcatecholamines, in contradistinction to the experienceof Roberts et al.7 and Wallach et al.,8 we and Pratilaset al.™ found that both hypertensive and normotensivepatients had equally activated sympathetic systems.

The most impressive difference between thehypertensive and normotensive patients in our study isrelated to preoperative propranolol administration. Inpatients who subsequently developed hypertension,chronic propranolol administration was more com-mon and the dosage higher. Very similar cir-cumstances during animal studies were observed byTyers and Hughes.23 In their experiments, bloodpressure at 1 hour following CBP was decreased incontrol dogs but significantly increased (p < 0.05) indogs treated with oral propranolol until 8 hours priorto surgery.

Hypertension in the setting of propranololwithdrawal could result either from excessive sym-pathetic stimulation in the presence of continued beta-receptor blockade or from the occurrence of arebound hyperreninemic state or hyperkinetic stateresulting from excessive beta-adrenergic activityfollowing abrupt cessation of therapy. Although oralpropranolol is usually said to have a relatively shorthalf-life," prolonged beta-blocking effects have beenreported following discontinuation of propranolol inboth animal28 and human"- M studies.

In our study, propranolol was generally discon-tinued 6 to 8 hours before surgery; therefore, it seemslikely that beta blockade could still have been presentin many of our patients. This is particularly true ofthe patients who developed hypertension since theyhad received the largest preoperative propranololdoses29- ^ and had a significantly lower heart rate thantheir normotensive counterparts prior to induction ofanesthesia. Propranolol, being a potent beta-adrenergic receptor antagonist, blocks the localvasodilating responses that accompany adrenergicstimulation. Thus, catecholamine release duringpropranolol therapy would produce unopposed alpha-adrenergic stimulation with resulting vasoconstrictionand hypertension. Hypertension that is reversible byalpha-adrenergic blockade has been noted duringpropranolol therapy of hyperadrenergic states'1' M andhas also been reported following catecholamine ad-ministration to both hypertensive" and normoten-sive84 patients receiving propranolol. A similarmechanism has been proposed to explain the paradox-ical hypertension occasionally reported during chronicpropranolol therapy." Since the hypertension noted inour study occurred in the setting of recent propranololtherapy, elevated levels of plasma catecholamines, andan increase in peripheral vascular resistance, it may becaused by unopposed alpha-adrenergic stimulation, atleast in part.

An alternative mechanism might be a rebound statefollowing the rapid tapering or discontinuing ofchronic propranolol therapy. Rebound hypertensionhas been reported following abrupt withdrawal ofsympatholytic agents."' " In this instance, however itseems unlikely, as one would expect such a state to bemanifested by an increase in circulating levels of Allor in CI resulting from beta-receptor hypersensitivityto adrenergic stimulation. In addition, propranololrebound has most frequently been reported 24 to 72hours following its discontinuation,**' ** considerablylater than the timing of the hypertensive episodesnoted in our study.

If hypertension following coronary artery bypasssurgery does result from unopposed alpha-adrenergicstimulation, one would expect sympathetic blockadeto be effective in lowering the patient's blood pressure,and indeed, this has been demonstrated with boththoracic epidural anesthesia6 and phentolamine ad-ministration.12 Although it is equally true that otheragents such as sodium nitroprusside4'7'10> ll and con-verting enzyme inhibitor1' are also capable of decreas-ing blood pressure in this situation, this does notnegate the role of the sympathetic system. Sodiumnitroprusside is known to be effective even in patientswith refractory hypertension resulting from pheo-chromocytoma,40 and converting enzyme inhibitor hasmany actions other than blockade of the renin-angiotensin system.41

It must be emphasized that our findings were ob-tained in a relatively small group of patients and thatthe importance of these observations can only be con-firmed by further studies. Clearly, propranolol is notthe only, nor perhaps the most important,pathogenetic factor involved. Some of our patientsdeveloped hypertension despite the fact that they hadnot received propranolol prior to surgery. In addition,Estafanous et al.4 noted hypertension following cor-onary artery bypass surgery in 35% of 76 patients whohad not received beta blockers for at least 48 hoursprior to surgery.

Our results do suggest, however, that preoperativepropranolol therapy may increase the risk of develop-ing hypertension following coronary artery bypasssurgery. Perhaps the rapid preoperative tapering ordiscontinuing of propranolol causes a rebound state,or, more likely, the propranolol-mediated beta-receptor blockade causes unopposed alpha-adrenergicstimulation, resulting in hypertension. Preoperativepropranolol therapy should not be abruptly withdrawnin any patient as this may precipitate the developmentof unstable angina, myocardial infarction, or evensudden death." It may not be possible to taperpropranolol therapy in some patients with severeangina. However, gradual tapering and discontinua-tion of propranolol several days prior to coronaryartery bypass surgery is possible in the majority ofpatients and might decrease the prevalence ofpostoperative hypertension. Naturally, the value ofgradual tapering as compared to rapid tapering or notapering at all can only be proven by a prospectiverandomized trial, which awaits further study.

by guest on July 19, 2018http://hyper.ahajournals.org/

Dow

nloaded from

298 HYPERTENSION VOL 2, No 3, MAY-JUNE 1980

References

1. Criteria Committee of the New York Heart Association:Diseases of the Heart and Blood Vessels (Nomenclature andCriteria for Diagnosis), 6th edition. Boston, Little, Brown andCompany, 1964, p 110

2. Walker WG, Horvath JS, Moore MA, Whelton PK, RussellRP: Relation between plasma renin activity, angiotensin andaldosterone and blood pressure in mild untreated hypertension.Circ Res 38: 470, 1976

3. Passon PG, Peuler JD: A simplified radiometric assay forplasma norepinephrine and epinephrine. Anal Biochem 51:618,1972

4. Estafanous FG, Tarazi RC, Viljoen JF, ElTawil MY: Systemichypertension following myocardial revascularization. AmHeart J 85: 732, 1973

5. Hoar PF, Hickey RF, Ullyot DJ: Systemic hypertensionfollowing myocardial revascularization. J Thorac CardiovascSurg 71: 859, 1976

6. Bailey DR, Miller ED, Kaplan JA, Rogers PW: The renin-angiotensin-aldosterone system during cardiac surgery withmorphine-nitrous oxide anesthesia. Anesthesiology 42: 538,1975

7. Roberts AJ, Niarchos AP, Subramanian VA, Abel RM, Her-man SD, Sealey JE, Case DM, White RP, Johnson GA,Laragh JM, Gay WA: Systemic hypertension associated withcoronary artery bypass surgery. J Thorac Cardiovasc Surg 74:846, 1977

8. Wallach R, Karp RB, Reves JG, Oparil S, James TN:Mechanism of hypertension after saphenous vein bypass sur-gery. Circulation 56 (suppl III): III-544, 1977

9. Haggendal E, Johansson B: Physiological aspects of the bloodbrain barrier change in acute arterial hypertension. Eur Neurol6: 24, 1972

10. Viljoen JF, Estafanous FG, Tarazi RC: Acute hypertension im-mediately after coronary artery surgery. J Thorac CardiovascSurg 71: 548, 1976

11. Fouad FM, Estafanous FG, Tarazi RC: Hemodynamics ofpostmyocardial revascularization hypertension. Am J Cardiol41: 564, 1978

12. Roberts AJ, Niarchos AP, Case DB, Abel RM, SubramanianVA, Laragh JH, Gay WA: Coronary artery bypass hyperten-sion: Comparison of responses to nitroprusside, phentolamineand converting enzyme inhibitor. Circulation 56 (suppl III):III-217, 1977

13. Niarchos AP, Roberts AJ, Case DB, Gay WA, Laragh JH:Hcmodynamic characteristics of hypertension after coronarybypass surgery and effects of the converting enzyme inhibitor.Am J Cardiol 43: 586, 1979

14. Tarazi RC, Estafanous FG, Fouad FM: Unilateral stellateblock in the treatment of hypertension after coronary bypasssurgery. Am J Cardiol 42: 1013, 1978

15. Hasbrouck JD: Morphine anesthesia for open-heart surgery.Ann Thorac Surg 10: 364, 1970

16. Motlagh F, Alavi F, Natmabadi MH, Sheikhzadem A, MoezziB: The relation of cardio-pulmonary bypass induced hyperten-sion and renin-angiotensin system. Circulation 56 (suppl III):III-546, 1977

17. James TN, James HI, Urthaler F: Analysis of components in acardiogenic hypertensive chemoreflex. Circulation 52: 179,1975

18. Eckstein RW, Shintani F, Rowen HE Jr, Shimomura K, OhyaN: Identification of left coronary blood supply of aortic bodiesin anesthetized dogs. J Appl Physiol 30: 488, 1971

19. Brown AM, Malliani A: Spinal sympathetic reflexes initiatedby coronary receptors. J Physiol 212: 685, 1971

20. Pratilas V, Vlachakis ND, Litwak R: Hypertension and plasmacatecholamines following aorto-coronary by-pass surgery. ClinRes 25: 244A, 1977 (abstr)

21. Replogle R, Levy M, DeWall RA, Lillehei RC: Catecholamineand serotonin response to cardiopulmonary bypass. J ThoracCardiovasc Surg 44: 638, 1962

22. Halter JB, Pflug AE, Porte D Jr: Mechanism of plasmacatecholamine increases during surgical stress in man. J ClinEndocrinol Metab 45: 936, 1977

23. Tyers GFO, Hughes HC: Circulatory effects of propranolol andcardiopulmonary bypass. J Surg Res 18: 181, 1975

24. Holland OB, Kaplan NM: Propranolol in the treatment ofhypertension. N Engl J Med 294: 930, 1976

25. Myers JH, Horwitz LD: Hemodynamic and metabolic responseafter abrupt withdrawal of long-term propranolol. Circulation58: 196, 1978

26. Boudoulas H, Dervenagas S, Lewis R, Kates R, DalamangasG: Time course of the blockade effect of propranolol on sinusnote and atrioventricular node. Clin Res 26: 589A, 1978 (abstr)

27. Paterson JW, Conolly ME, Dollery CT, Hayes A, Cooper RG:The pharmacodynamics and metabolism of propranolol inman. Pharmacol Clinica 2: 127, 1970

28. Shiroff RA, Levenson LW, Zelis R, Shiroff JE, Schneck DW,Hayes AH: Duration of action of propranolol on heart rate dur-ing exercise stress testing. Cardiovasc Med 3: 1029, 1978

29. Conolly ME, Kersting F, Dollery CT: The clinical phar-macology of beta-adrenoceptor-blocking drugs. Prog Car-diovasc Dis 19: 203, 1976

30. Romagnoli A, Keats AS: Plasma and atrial propranolol afterpreoperative withdrawal. Circulation 52: 1123, 1975

31. Prichard BNC, Ross EJ: Use of propranolol in conjunctionwith alpha receptor blocking drugs in pheochromocytoma. AmJ Cardiol 18: 394, 1966

32. Blum I, Atsmon A, Steiner M, Wysenbeek H: Paradoxical risein blood pressure during propranolol treatment. Br Med J 4:623, 1975

33. Vlachakis ND, DcGuia D, Mendlowitz M: Blood pressureresponses to catecholamines during beta-adrenergic blockadewith propranolol in hypertensive subjects. Chest 71: 38, 1977

34. Harris WS, Schoenfeld CD, Brooks RH, Weissler AM: Effectof beta-adrenergic blockade on the hemodynamic responses toepinephrine in man. Am J Cardiol 17: 484, 1966

35. McMurtry RJ: Propranolol, hypoglycemia and hypertensivecrisis. Ann Intern Med 80: 669, 1974

36. Hunyor SN, Hansson L, Harrison TS, Hoobler SW: Effects ofclonidine withdrawal: Possible mechanisms and suggestions formanagement. Br Med J 2: 209, 1973

37. Frewin DB, Penhall RK: Rebound hypertension after suddendiscontinuation of methyl dopa therapy. Med J Aust 1: 659,1977

38. Boudoulas H, Lewis RP, Kates RE, Dalamanagas G:Hypersensitivity to adrenergic stimulation after propranololwithdrawal in normal subjects. Ann Intern Med 87: 433, 1977

39. Nattel S, Rangno RE, VanLoon GR: Mechanism ofpropranolol withdrawal phenomena. Clin Res 26: 292A, 1978(abstr)

40. Lipson A, Hsu TH, Sherwin B, Geelhoed FW: Nitroprussidetherapy for a patient with a pheochromocytoma. JAMA 239:427, 1978

41. Williams GH: Angiotensin-dependent hypertension-potentialpitfalls in definition. N Engl J Med 296: 684, 1977

42. Shand DG, Wood AJJ: Propranolol withdrawal syndrome —why? Circulation 58: 202, 1978

by guest on July 19, 2018http://hyper.ahajournals.org/

Dow

nloaded from

G WalkerP K Whelton, J T Flaherty, N P MacAllister, L Watkins, A Potter, D Johnson, R P Russell and W

therapy.Hypertension following coronary artery bypass surgery. Role of preoperative propranolol

Print ISSN: 0194-911X. Online ISSN: 1524-4563 Copyright © 1980 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Hypertension doi: 10.1161/01.HYP.2.3.291

1980;2:291-298Hypertension. 

http://hyper.ahajournals.org/content/2/3/291World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://hyper.ahajournals.org//subscriptions/

is online at: Hypertension Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer process is available in the

Request Permissions in the middle column of the Web page under Services. Further information about thisOffice. Once the online version of the published article for which permission is being requested is located, click

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialHypertension Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on July 19, 2018http://hyper.ahajournals.org/

Dow

nloaded from